All MCP servers
Browser & webOfficial
Firecrawl
By Mendable
When fetch isn't enough because the content is behind JS or you need a whole sub-tree of pages, Firecrawl is the standard answer.
What you can do with it
- Pull every page of a docs site into a single context dump.
- Scrape a SPA that fetch can't see.
- Build a one-off knowledge base from a competitor's help center.
Works with
Claude DesktopClaude CodeCursorZedWindsurfCline
Other browser & web servers
Fetch
Download a URL and return clean markdown the model can read.
Brave Search
Search the web and get back results with summaries and citations.
Tavily
Search optimised for AI agents — returns clean, ranked, citation-ready content.
Playwright
Drive a real browser — click, type, fill forms, take screenshots.